Speech recognition technology is becoming increasingly popular in a variety of computerized devices. One use of this technology is to offer real-time closed captioning to individuals with hearing difficulties where speech is captured by a microphone, and then real-time transcription is performed by a computer. In group conversations, however, problems arise in applying speech recognition to members of the group. Unless a speaker is very close to the microphone, the resulting audio signal is likely to be of poor quality which causes the speech recognition to be correspondingly poor.